456 PROCEEDINGS OF THE AMERICAN ACADEMY 



The writer has described (loc. cit.) the remarkable hygroscopic pro- 

 perties of the vermiculites, and the difficulty of separating the consti- 

 tutional from the hygroscopic water which they may contain. The 

 varieties from Lerni and Pelham offer the same difficulty in the 

 determination of their water, thirty to forty hours being required to 

 brintj one or two ijrammes of either of them to a constant weight at 



100* c. 



In obtaining a constant temperature of 100° C, an electric regidator 

 was used which diffisrs from other similar forms of apparatus in simplic- 

 ity of construction. The current is made or broken by a very slight 

 rise or fall of mercury in a U tube connected with a glass bulb within 

 the air-bath. By means of a pressure tap which closes an open L of 

 the connecting tube, the air within the bulb may be confined as soon 

 as the bath reaches the required temperature. After this a very slight 

 increase of temperature raises the mercury column sufficiently to close 

 the electric circuit, and then the current shuts the cock which regu- 

 lates the sup[)ly of gas to the burner under the bath. The chief ad- 

 vantage and tiie novelty of the apparatus is to be found in the 

 simplicity of this stop-cock, which was suggested by Professor H. B. 

 Hill. It consists of an ordinary chloride of calcium tube placed hori- 

 zontally, and closed at the larger end by a rubber stopper which allows 

 a considerable freedom of motion to a smaller glass tube passing 

 through it ; by this the illuminating gas enters the chloride of calcium 

 tube, from which it passes to the burner. When the current is closed, 

 an electro-mairnet acting on an armature attached to the outer end of 

 the small tube plunges the curved inner end beneath the surface of 

 some mercury in the bulb of the chloride of calcium tube, and thus 

 shuts off the main supply of gas ; although a small orifice in the side 

 of the inner tube allows a sufficient flow to keep the flame under the air- 

 bath alive. The variation of the temperature of the air-bath does not 

 ordinarily exceed one or two degrees during periods of fifteen to twenty 

 hours, even under great variations of pressure in the gas mains. 



Table I. shows the percentages of water found in air-dried specimens 



